How likely would this still have happened had the Dod-Frank regulations applicable to banks this size had not been significantly loosened back around 2018 (if I recall correctly)? Thanks for the great analysis.
My understanding is they would have been made less likely, particularly by the fact that SVB's deposits would have correctly been assessed as riskier than most banks, and I think we will see a regulatory clampdown on regional banks in the wake of this.
Excellent description. This is banking 101. But it is far from clear that it is politically feasible for the Fed to bailout some of the wealthiest people on the planet. It IS clear that monetary policy is in the frame as a proximate cause. It is also clear that policymakers started the year looking to starve crypto companies of liquidity. Had SVB followed basic ALM principles, it would not be in this position. So this must also be a story of regulatory failure.
Great overview! It's starting to become an interesting situation with banking - similar to what some people say happens in aviation. With all the automation in aviation, some have argued that pilots are now worse and this puts planes in danger (whether that's true is up for debate). With these banking decisions, it fees like just because the regulations allowed for something, SVB management made a decision to do it, by holding a large amount of assets, bought at peak prices without any interest rate hedges. It feels like just because they weren't told not to do that, they didn't. It might not be easy to come up with regulations for every eventuality. It makes me wonder if the decision makers (CFOs, operational finance etc) should be required to get certifications so that we're sure that they actually understand the risks of their business. Especially since the banking sector is treated very differently from all other sectors.
Another sad and detailed expose of our banking system's foibles. How do seemingly professionals make such stupid and fool-hearty mistakes? Not encouraging but appreciate the insight into the follies of these mega-institutions. Once again, who is watching the store??
“As of the end of 2023, the fair market value of the bank’s total held-to-maturity securities portfolio was more than $15B smaller than the assets’ amortized value, a number that had only grown larger and larger as interest rates increased.”
Today, he says, “we are in the golden age of fraud”.
Chanos describes the current environment as “a really fertile field for people to play fast and loose with the truth, and for corporate wrongdoers to get away with it for a long time”. He reels off why: a 10-year bull market driven by central bank intervention; a level of retail participation in the markets reminiscent of the end of the dotcom boom; Trumpian “post-truth in politics, where my facts are your fake news”; and Silicon Valley’s “fake it until you make it” culture, which is compounded by Fomo — the fear of missing out. All of this is exacerbated by lax oversight. Financial regulators and law enforcement, he says, “are the financial archaeologists — they will tell you after the company has collapsed what the problem was.”
All in all, it’s “a heady witch’s brew for trouble”.
Is there a specific link to the data that you used for the Uninsured and Insured Deposits graph? i tried searching for a while but couldn't get anywhere
I have the same question. (It could be simply that they went longer to try and squeeze out a bit more yield, but perhaps there’s more to it than that.)
SVB had a lot of deposit growth (both from an increase in national monetary base and from a big inflow of transfers as venture capital raised a lot of money in 2021/2022) and had to match that deposit growth with something on the asset side. Normal banks are usually constantly making loans (indeed, in aggregate the loans are where the deposits come from) and have a lot more flexibility to change their asset profile—lend to businesses and households with lower credit, extend loan maturities, etc.
But if you're a tech bank and so many of your clients are fresh-faced startups who just raised millions in equity, who are you going to lend to? The startups don't want to borrow money, they just raised a ton of it. You can do boutique things like lend against startup equity or do venture debt projects but broadly speaking there's a cap on how much your clients want to borrow and its far less than what they now have in deposits.
So you go out and invest in a bunch of high quality assets with longer maturities. Treasuries and MBS have very low default risk and you squeeze out a bit more yield by holding them. It really only goes against you if rates go up rapidly, and even then you can theoretically hold to maturity without a problem. You take on duration risk instead of credit risk, and a lot of it, to make money.
Part of the core problem was the deposit base—these were not people who wanted traditional banking services, which is why they didn't bank with a traditional bank. They moved in quickly and moved out quickly, and SVB wasn't prepared for that. I am not defending the decision to invest in so many long-term assets at generationally low yields (it was a losing idea, but more importantly it was a risky idea that went unhedged and they paid dearly for that) but that was the logic.
> As of the end of 2023
Guessing this should say 2022?
yep haha! Fixed it
Excellent analysis
AFS is a “Banking Book” classification and does not require MTM treatment. Otherwise solid recap!
How likely would this still have happened had the Dod-Frank regulations applicable to banks this size had not been significantly loosened back around 2018 (if I recall correctly)? Thanks for the great analysis.
My understanding is they would have been made less likely, particularly by the fact that SVB's deposits would have correctly been assessed as riskier than most banks, and I think we will see a regulatory clampdown on regional banks in the wake of this.
Thank you.
Excellent description. This is banking 101. But it is far from clear that it is politically feasible for the Fed to bailout some of the wealthiest people on the planet. It IS clear that monetary policy is in the frame as a proximate cause. It is also clear that policymakers started the year looking to starve crypto companies of liquidity. Had SVB followed basic ALM principles, it would not be in this position. So this must also be a story of regulatory failure.
Great overview! It's starting to become an interesting situation with banking - similar to what some people say happens in aviation. With all the automation in aviation, some have argued that pilots are now worse and this puts planes in danger (whether that's true is up for debate). With these banking decisions, it fees like just because the regulations allowed for something, SVB management made a decision to do it, by holding a large amount of assets, bought at peak prices without any interest rate hedges. It feels like just because they weren't told not to do that, they didn't. It might not be easy to come up with regulations for every eventuality. It makes me wonder if the decision makers (CFOs, operational finance etc) should be required to get certifications so that we're sure that they actually understand the risks of their business. Especially since the banking sector is treated very differently from all other sectors.
Another sad and detailed expose of our banking system's foibles. How do seemingly professionals make such stupid and fool-hearty mistakes? Not encouraging but appreciate the insight into the follies of these mega-institutions. Once again, who is watching the store??
You mean at the end of 2022 right?
“As of the end of 2023, the fair market value of the bank’s total held-to-maturity securities portfolio was more than $15B smaller than the assets’ amortized value, a number that had only grown larger and larger as interest rates increased.”
Yep—fixed the error!
Today, he says, “we are in the golden age of fraud”.
Chanos describes the current environment as “a really fertile field for people to play fast and loose with the truth, and for corporate wrongdoers to get away with it for a long time”. He reels off why: a 10-year bull market driven by central bank intervention; a level of retail participation in the markets reminiscent of the end of the dotcom boom; Trumpian “post-truth in politics, where my facts are your fake news”; and Silicon Valley’s “fake it until you make it” culture, which is compounded by Fomo — the fear of missing out. All of this is exacerbated by lax oversight. Financial regulators and law enforcement, he says, “are the financial archaeologists — they will tell you after the company has collapsed what the problem was.”
All in all, it’s “a heady witch’s brew for trouble”.
--Jim Chanos...Enuf Said!
Bubbles do tend to burst.
Thank you for this insightful explainer, Joey! I may be an economist, but finance is undecipherable to me. Not anymore!
Is there a specific link to the data that you used for the Uninsured and Insured Deposits graph? i tried searching for a while but couldn't get anywhere
Hey Mart8! The data comes from FDIC Call Reports, you can find that for SVB and other banks here: https://banks.data.fdic.gov/docs/
Thank you very much!
I have the same question. (It could be simply that they went longer to try and squeeze out a bit more yield, but perhaps there’s more to it than that.)
SVB had a lot of deposit growth (both from an increase in national monetary base and from a big inflow of transfers as venture capital raised a lot of money in 2021/2022) and had to match that deposit growth with something on the asset side. Normal banks are usually constantly making loans (indeed, in aggregate the loans are where the deposits come from) and have a lot more flexibility to change their asset profile—lend to businesses and households with lower credit, extend loan maturities, etc.
But if you're a tech bank and so many of your clients are fresh-faced startups who just raised millions in equity, who are you going to lend to? The startups don't want to borrow money, they just raised a ton of it. You can do boutique things like lend against startup equity or do venture debt projects but broadly speaking there's a cap on how much your clients want to borrow and its far less than what they now have in deposits.
So you go out and invest in a bunch of high quality assets with longer maturities. Treasuries and MBS have very low default risk and you squeeze out a bit more yield by holding them. It really only goes against you if rates go up rapidly, and even then you can theoretically hold to maturity without a problem. You take on duration risk instead of credit risk, and a lot of it, to make money.
Part of the core problem was the deposit base—these were not people who wanted traditional banking services, which is why they didn't bank with a traditional bank. They moved in quickly and moved out quickly, and SVB wasn't prepared for that. I am not defending the decision to invest in so many long-term assets at generationally low yields (it was a losing idea, but more importantly it was a risky idea that went unhedged and they paid dearly for that) but that was the logic.